You did, though:
A video image is made up of pixels, but it's drawn on the screen in horizontal lines. The number refers to the number of lines.
What makes a high quality image? The technical aspect of Image quality in Cameras
- Thread starter Ebrahim Saadawi
- Start date
You did, though:
A video image is made up of pixels, but it's drawn on the screen in horizontal lines. The number refers to the number of lines.
Ebrahim Saadawi
Carbonite Member
The P "number" refers to the number of pixels in a row, is not the same as the letter "p" stands for "progressive, or pixels''
Anyway English is not my first language, in fact, it's quite limited. So I guess some of my choice of words can confuse some people. I'll change that part to a more clear sentense. Thank you both for the note.
But about the lines vs pixels. Are we sure the 1080p number refers to the number of lines not the number of pixels present in one row? When we speak of photography a 2mp photo is 1920x1080 means 1920 pixels in the first horizontal row and 1080p pixels in the first vertical row, when both multiplied, it gives us the amount of all the pixels in between (around 2 million). I thought it's how these numbers come from. Is that incorrect?
We're sure. 
j1clark@ucsd.edu
Veteran
The Daughter asked me about the difference between progressive and interlace 2 weeks ago when she was looking around for a camera that would do reasonable motion picture filming for her 'minimal' budget. Since the Family is in the Nikon camp, and she has one of our older still cameras, she was looking at the Nikon D5200 vs D5300... or something in that category...
So, yeah, I think some mention of the difference, and the 'advice' of 'use progressive' would be good.
The 1080/720 business refers to 'video lines' and a hold over from Real(tm) Analog Video days. Analog video cameras were spec'd with how many vertical scan lines. For the US, some portions of the Americas, and Japan, the NTSC standard was 525 scan lines, but 485 were 'active'... DV sort compromised on 480... close enough for government work... Analog video signals also used some of those 'non displayed' scan lines to encode 'caption' and other information... PAL and SECAM standards covered the rest of the world, but were similar in nature to NTSC. Things were further confused with 'square' vs. 'rectangular' pixels, and so one would see 640 horizontal pixes for NTSC, vs 720 (not to be confused with the HD standard 720p...), the difference being that one had 640 square pixels or 720 rectangular pixels...
This is subtle, but when discussion Digital Film vs Video, one sees for Digital film such specs as 2K, 2.5K, 4K, etc, referring to the horizontal pixel counts... when discussing Digital video one sees 1080(i,p), 720(i,p), 480(i,p)... that is the latter is allowing the person who has experience in 'video' to feel more comfortable. Where as Digital film is in horizontal pixels, and was used for 'scanning' Film film prior to the development of original capture devices such as Red and Alexas... When one goes to the Sony site, and looks at their 'pro' line, the specs have both styles of 'resolution' listed (along with dB and ISO values for exposure settings...).
Last edited:
combatentropy
Veteran
It doesn't help that things like "2K," "4K," and "UHD" are mostly marketing terms rather than anything actually technical. 1920x1080 is basically 2K (and considered within the "2K family"), but "2K" (two thousand!) sounds more impressive than "1080." Even worse, when making the (again, marketing!) comparison of "4K" to "1080," there's a sleight-of-hand involved. Yes, 4K has "four times the resolution" of 1080, but that's not because 4K, or 4,000, is four times 1,000.
It also doesn't help that "p" and "i" can be attached to frame rate as well as resolution, so someone ENTIRELY uninitiated may look at "1080p" and "24p" and get entirely the wrong idea about one of them, or even both.
There are a lot of legacy, throwback terms which need to be straightened out and made consistent. The inaccurate and misleading designations of "30i" and "29.97i" and "25i" need to go away, as I argue in 24p & Frame Rates; the nomenclature of resolution needs to be standardized, etc. Problem is, of course, that there is no standardization body for these things . . .
(On the "29.97" thing -- the Mythbusters referred to it in the latest "Indiana Jones" episode, and Adam used it in a way which confused the issue.)
Ryan Patrick O'Hara
Veteran
If you wanted to keep it simple enough for beginners I would simply state the primary factor that determines the shape and quality of bokeh is the number and shape of the aperture blades in the iris assembly.(An iris assembly with 5 curved blades can have a rounder shape than a lens with 7-9 straight edge blades).btw
I would then give examples and move on. No need to pass judgement on one or the other. If this were about aesthetics, than maybe that would be appropriate, but it seems you are trying to be technical and observant, so perhaps just an explanation is necessary.
Also, you refer to your example image as being hexagonal. It is not. It is pentagonal in shape. Furthermore, iris assemblies come in all different designs, so it's often best to just refer to any non-round iris as polygonal. That covers all bases.
Regarding anamorphic bokeh, an anamorphic lens is comprised of two basic components. A spherical taking lens and an anamorphic attachment, either front or rear. The front anamorphic element is the one that will give you the strong anamorphic characteristics, one of which is the oval bokeh. In this instance, the odd oval shape of the bokeh is not the result of the shape or number of the iris blades.
Patryk isn't bad company
sonofaresiii
Well-known member
What's the difference? It's a grid.
And also, I think you're misunderstanding what most people refer to when they say barrel/pincushion distortion on lenses. It's not the distortion that comes from a perspective change-- many lenses have the distortion present regardless of your perspective to the subject.
sonofaresiii
Well-known member
ps nice write-up overall
j1clark@ucsd.edu
Veteran
Hopefully the use of TV Lines for resolution has been all but forgotten by most, other than historians...
But I do think of "K" in two ways... in regard to things computer... K = 1024... everything else "K" = 1000...
And when confronted with 'odd' dimensions, even with "K" = 1024... I start thinking about bus widths and cache line widths...
Ryan Patrick O'Hara
Veteran
Here on the left a Fullframe sensor and on the right an APS-C smaller sensor at the settings/lighting
^^^Where did you get this example? From my experience it's highly misleading.
And again, larger sensors don't make cleaner images.... larger photosites do, and even that is a dangerous summarization to make.
For instance, a Sony A7R and a Sony A7S are very similar cameras. Both are Full Frame. However, one is 36.4mp and the other is 12mp. The Sony A7S is much more low-light friendly because it has larger photo-sites, not a larger sensor.
Larger sensors have the potential for great low-light, but size of sensor does not guarantee performance. In addition, photosite size isn't the only factor that determines the sensitivity of a camera.
Last edited:
Ebrahim Saadawi
Carbonite Member
When I determine how a factor has an effect on IQ I assume my readers understand it's when all the other elements are kept constant.
For example the image I provided is a comparison in high ISO performance between a 550D, and a 1Dx at 6400 ISO. Because both of then have the exact same pixel count at 18mp, both the sane technology generation from the same manufacturer, and in the picture shot at the same settings and conditions, but the 1Dx is fullframe and the 550D is APS-C.
So I think it's useful for beginners to have the impression of why larger full frame sensors are desired and more expensive when they read the price difference between a dx vs fx or APS-C vs fullframe (the immediate impression being better lowlight performance usually and shallower dof,
If we start to introduce other variables like sensor technology, pixel count and size, noise reduction processing, then a smaller sensor with these advantages could outperform a larger sensor in lowlight. So I do get your point that it's not a necessaty for a larger sensor to bet a smaller one in lowlight due to other factors.
Should I also add all the other factors that effect lowlight performance or would it complicate the sensor paragraph for beginners?
For example the image I provided is a comparison in high ISO performance between a 550D, and a 1Dx at 6400 ISO. Because both of then have the exact same pixel count at 18mp, both the sane technology generation from the same manufacturer, and in the picture shot at the same settings and conditions, but the 1Dx is fullframe and the 550D is APS-C.
So I think it's useful for beginners to have the impression of why larger full frame sensors are desired and more expensive when they read the price difference between a dx vs fx or APS-C vs fullframe (the immediate impression being better lowlight performance usually and shallower dof,
If we start to introduce other variables like sensor technology, pixel count and size, noise reduction processing, then a smaller sensor with these advantages could outperform a larger sensor in lowlight. So I do get your point that it's not a necessaty for a larger sensor to bet a smaller one in lowlight due to other factors.
Should I also add all the other factors that effect lowlight performance or would it complicate the sensor paragraph for beginners?
Samuel H
Major Contributor
Let me say again that this will be a great read-up for newbies.
Also, I aggree that as a general rule bigger sensors will give you sharper images with less noise. But it is indeed a rule with lots and lots of exceptions, because tech matters more than sensor size (so, small sensor with better tech, say GH4, can be a lot sharper than bigger sensor with bad tech, say 5D3+H.264; at same tech -5D2 vs T2i- it usually holds).
Also, I aggree that as a general rule bigger sensors will give you sharper images with less noise. But it is indeed a rule with lots and lots of exceptions, because tech matters more than sensor size (so, small sensor with better tech, say GH4, can be a lot sharper than bigger sensor with bad tech, say 5D3+H.264; at same tech -5D2 vs T2i- it usually holds).
combatentropy
Veteran
@combatentropy : again, good morning mate
Good afternoon, greymog. Hope you're doing well.
Ryan Patrick O'Hara
Veteran
To simply follow up, here are two iris assemblies at T/2.8
I just happened to have both makes of lenses on a current job and I thought about this thread.
These are arguably the two most popular non-panavision brand cinema lens sets for the last decade or so.
Both are amazing lens designs and sets. One has 8-blades, and the other has 9-blades. You can see how different the shapes can be, despite similar blade count. Iris blade shape is equal to importance, if not more important, than number of blades.
I just happened to have both makes of lenses on a current job and I thought about this thread.
These are arguably the two most popular non-panavision brand cinema lens sets for the last decade or so.
Both are amazing lens designs and sets. One has 8-blades, and the other has 9-blades. You can see how different the shapes can be, despite similar blade count. Iris blade shape is equal to importance, if not more important, than number of blades.
Last edited:
